Article PDF - Christopher A. Walsh Laboratory

Am. J. Hum. Genet. 71:1033–1043, 2002
Mutations in the O-Mannosyltransferase Gene POMT1 Give Rise
to the Severe Neuronal Migration Disorder Walker-Warburg Syndrome
Daniel Beltrán-Valero de Bernabé,1 Sophie Currier,3 Alice Steinbrecher,4,5 Jacopo Celli,1
Ellen van Beusekom,1 Bert van der Zwaag,2 Hülya Kayserili,6 Luciano Merlini,7
David Chitayat,8 William B. Dobyns,9 Bru Cormand,10 Ana-Elina Lehesjoki,11 Jesús Cruces,12
Thomas Voit,5 Christopher A. Walsh,3 Hans van Bokhoven,1 and Han G. Brunner1
Departments of 1Human Genetics and 2Pathology, University Medical Centre Nijmegen, Nijmegen, The Netherlands; 3Division of Neurogenetics,
Beth Israel Deaconess Medical Center and Harvard Medical School, Boston; 4Institute of Neuropathology, Free University Berlin, Berlin;
5
Department of Pediatrics and Pediatric Neurology, University Hospital Essen, Essen; 6Department of Neurology, Division of Medical
Genetics, Istanbul Medical Faculty, Istanbul University, Istanbul; 7Laboratory of Neuromuscular Pathology, Istituto Ortopedico Rizzoli,
Bologna; 8Department of Obstetrics and Gynecology, Mount Sinai Hospital, The University of Toronto, Toronto; 9Departments of Human
Genetics, Neurology, and Pediatrics, The University of Chicago, Chicago; 10Department of Genetics, University of Barcelona, Barcelona;
11
Folkhalsan Institute of Genetics and Department of Medical Genetics, University of Helsinki, Helsinki; and 12Department of Biochemistry,
Faculty of Medicine, University Autónoma of Madrid, Madrid
Walker-Warburg syndrome (WWS) is an autosomal recessive developmental disorder characterized by congenital
muscular dystrophy and complex brain and eye abnormalities. A similar combination of symptoms is presented
by two other human diseases, muscle-eye-brain disease (MEB) and Fukuyama congenital muscular dystrophy
(FCMD). Although the genes underlying FCMD (Fukutin) and MEB (POMGnT1) have been cloned, loci for WWS
have remained elusive. The protein products of POMGnT1 and Fukutin have both been implicated in protein
glycosylation. To unravel the genetic basis of WWS, we first performed a genomewide linkage analysis in 10
consanguineous families with WWS. The results indicated the existence of at least three WWS loci. Subsequently,
we adopted a candidate-gene approach in combination with homozygosity mapping in 15 consanguineous families
with WWS. Candidate genes were selected on the basis of the role of the FCMD and MEB genes. Since POMGnT1
encodes an O-mannoside N-acetylglucosaminyltransferase, we analyzed the possible implication of O-mannosyl
glycan synthesis in WWS. Analysis of the locus for O-mannosyltransferase 1 (POMT1) revealed homozygosity in
5 of 15 families. Sequencing of the POMT1 gene revealed mutations in 6 of the 30 unrelated patients with WWS.
Of the five mutations identified, two are nonsense mutations, two are frameshift mutations, and one is a missense
mutation. Immunohistochemical analysis of muscle from patients with POMT1 mutations corroborated the Omannosylation defect, as judged by the absence of glycosylation of a-dystroglycan. The implication of O-mannosylation in MEB and WWS suggests new lines of study in understanding the molecular basis of neuronal migration.
Introduction
Neuronal migration is a complex process by which the
postmitotic neurons generated in the ventricular zone
disperse and are positioned into six distinct layers in
the mammalian neocortex. This process takes place
during the 3rd and 4th mo of gestation and is characterized by an inside-out migration pattern, by which
the successively newer neurons bypass the older ones
to occupy a more exterior layer in the brain cortex
(Marin-Padilla 1978). More than 25 neuronal migration disorders, resulting in death or improper posiReceived June 3, 2002; accepted for publication July 16, 2002;
electronically published October 4, 2002.
Address for correspondence and reprints: Dr. Han G. Brunner, 417
Department of Human Genetics, P.O. Box 9101, 6500 HB Nijmegen,
The Netherlands. E-mail: [email protected]
䉷 2002 by The American Society of Human Genetics. All rights reserved.
0002-9297/2002/7104-0004$15.00
tioning of the cortical neurons, have been described
in humans (Lammens 2000). The deficient neuronal
migration is accompanied by the absence of gyration
of the brain surface, referred to as “smooth brain” or
“lissencephaly.”
Walker-Warburg syndrome (WWS [MIM 236670])
is a recessive disorder characterized by severe brain
malformations, muscular dystrophy, and structural eye
abnormalities. The brain manifests cobblestone lissencephaly with agenesis of the corpus callosum, cerebellar hypoplasia, hydrocephaly, and sometimes encephalocele (fig. 1). Cobblestone lissencephaly, also
known as “cobblestone complex,” is caused by neural
overmigration, during neocortex lamination, that gives
rise to disorganized cerebral and cerebellar cortexes
and multiple coarse gyri with agyric regions. Two other
human syndromes present with a similar association
of symptoms: Fukuyama congenital muscular dystrophy (FCMD [MIM 253800]) and muscle-eye-brain dis1033
1034
Am. J. Hum. Genet. 71:1033–1043, 2002
Figure 1
Brain abnormalities in patients with WWS who have POMT1 mutations. A, Normal brain of fetus at 21 wk gestation. B, Brain
from an affected fetus from family 4 at 19 wk gestation. Note the rough cobblestone brain surface, the cerebellar hypoplasia (arrow), and the
abnormal vasculature. C, Midsagital T1–weighted magnetic-resonance image from an affected child from family 2, showing enlarged ventricles,
lack of corpus callosum, absence of pons, and rudimentary cerebellar vermis. D–F, Haematoxylin and eosin staining of paraffin-embedded brain
sections. D, Fetal brain at 21 wk gestation, showing normal lamination of the cortex: meninges (M), lamina molecularis (I), lamina granularis
externa (II), lamina pyramidalis (III), lamina granularis interna (IV), lamina ganglionaris (V), lamina multiformis (VI), and intermediate zone
(IZ). E, Brain section from affected fetus from family 4, showing abnormal cortical lamination with disruption of the glia limitans (**), and
protrusion of the migrating neuroblasts into the subarachnoid space (*). SZ p subventricular zone. F, Higher magnification of panel E, showing
the subarachnoid-space invasion by glio-vascular neuronal tissue. Glia limitans is shown with an arrow.
ease (MEB [MIM 253280]). Clinical and genetic analysis has revealed that WWS represents an independent
entity (Dobyns et al. 1989; Cormand et al. 2001).
WWS is the most severe syndrome of the group, especially with regard to the brain phenotype. Life expectancy of patients with WWS is usually !1 year. In
addition, recent linkage studies showed that WWS is
not allelic to the other two syndromes (Cormand et
al. 2001), but the loci for WWS have remained elusive.
Protein glycosylation is a highly complex mechanism
by which sugars are sequentially added to proteins at
the endoplasmic reticulum and the Golgi apparatus.
This posttranslational process modulates protein stability, conformation, and function and has been implicated in diverse molecular recognition events, such
as cell adhesion, growth, and differentiation (Kobata
1992; Varki 1993). The protein-attached glycans are
divided into two groups on the basis of their linkage
site: the N-glycans are linked to an asparagine residue
of the target protein, whereas the O-glycans are attached through a serine or a threonine. Several Olinked glycans can be distinguished by the sugar moiety
that O-links with the protein—for example, mannose,
N-acetylgalactosamine, or fucose (Endo 1999). O-
Beltrán-Valero de Bernabé et al.: POMT1 Mutations in WWS
mannosyl glycan synthesis is well studied in yeast,
but little is known about this process in higher eukaryotes (Strahl-Bolsinger et al. 1999). Among the
few O-mannosylated proteins that have been identified in mammals are a-dystroglycan, N-CAM, and
tenascin-J1 (Wing et al. 1992; Chiba et al. 1997; Sasaki et al. 1998). All of the O-mannosyl glycans sequences identified, although diverse, share the common motif galactose-b-1, 4-N-acetylglucosamine-b-1,
2-mannose-O-Ser/Thr (Galb1 4GlcNAcb1 2Man-OSer/Thr). So far, O-mannose–linked glycosylation has
been observed only in brain, peripheral-nerve, and
muscle glycoproteins (Wing et al. 1992; Chiba et al.
1997; Yuen et al. 1997; Sasaki et al. 1998; Smalheiser
et al. 1998; Chai et al. 1999). It has been estimated
that 30% of all O-linked sugar chains in brain are
O-mannose based (Chai et al. 1999).
Recently, the causative genes for FCMD (Fukutin)
(Kobayashi et al. 1998) and MEB (POMGnT1) (Yoshida et al. 2001) have been cloned. The MEB gene
product, POMGnT1, adds an N-acetylglucosamine
residue to a preexisting protein O-linked mannose
(Yoshida et al. 2001). The function of Fukutin is unknown, but it has been hypothesized to be a glycosyltransferase, based on amino acid homology with
several phosphoryl-ligand transferases (Aravind and
Koonin 1999). In addition, the highly homologous
Fukutin-related protein (FKRP) is also a putative glycosyltransferase (Brockington et al. 2001a). Mutations
in FKRP give rise to congenital muscular dystrophy
with or without neural involvement (Brockington et
al. 2001a, 2001b). Hypoglycosylation of a-dystroglycan has been observed in muscle tissue from patients with mutations in POMGnT1, FCMD, and
FKRP. These observations indicate the implication of
a common and crucial glycan synthesis process in
these disorders.
In the present article, we describe that, in a proportion of patients with WWS, the disease is caused
by mutations in the POMT1 gene. This gene encodes
O-mannosyltransferase 1, the enzyme that putatively
catalyzes the first step in O-mannosyl glycan synthesis.
These results demonstrate a key function of O-mannose–linked glycosylation in neural migration and open
new lines of study in understanding the molecular basis
of brain and muscle development.
Subjects, Material, and Methods
Subjects
All index patients (n p 30) were diagnosed through
the documentation of cobblestone lissencephaly observed on brain imaging or postmortem investigation,
as described elsewhere (Cormand et al. 2001). In ad-
1035
dition, at least one of the following was present: eye
malformations, congenital muscular dystrophy or elevated creatine kinase (CK), or occipital encephalocele.
Affected siblings (including aborted fetuses) were diagnosed to be also affected if the diagnosis in the index
case was secure and if at least one of the typical features
was documented. The phenotypes of patients that had
POMT1 mutations are described in more detail below.
The parents of family 1 are first cousins of Turkish
origin and have been described elsewhere (family 6 in
Cormand et al. 2001). After three spontaneous abortions, a male sibling was born presenting with severe
hydrocephalus with dilatation of the third and fourth
ventricle and minimal cortical development, no gyri
visible, bifid cerebellum, and hypoplasia of the vermis
and of the cerebellar hemispheres. A cerebellar cyst was
observed. The corpus callosum appeared to be present.
Microphthalmia of left eye and exophthalmia of right
eye were noted. Hypoplastic genitalia were also present. Serum CK levels were highly elevated at more than
2,000 U/liter. The patient died at age 7 mo, and no DNA
is available. A second child (an affected female sibling)
died 15 min after birth. She presented with severe hydrocephaly, encephalocele, and bilateral cleft lip. DNA
from this patient was used for genetic analysis.
Family 2 is a consanguineous family (first cousins) of
Turkish origin living in Germany and has been described
elsewhere (family 14 in Cormand et al. 2001). There is
one healthy child, and there are three siblings diagnosed
with WWS, a girl who died at age 3 years and two fetuses.
The deceased girl had cobblestone lissencephaly, microphthalmia, buphthalmos, megalocornea, glaucoma, and
retinal dysplasia. Serum CK was 1,700 U/liter. One fetus
had an encephalocele. The brain of the other fetus is
shown in figure 1.
Family 3 is also a consanguineous Turkish couple (first
cousins) with one affected girl and two unaffected boys
and has been described elsewhere (family 8 in Cormand
et al. 2001). The girl died at age 2 mo. She presented
with severe hydrocephalic dilatation of the lateral and
third ventricles, hypoplasia of vermis and cerebellum,
cyst formation in the posterior fossa, and a DandyWalker–like deformation of the fourth ventricles and
brainstem. Eye malformations included bilateral buphthalmos, bilateral glaucoma, and hypertelorism. Serum
CK levels were higher than 2,000 U/liter.
Family 4 is a Gypsy family from Italy and has been
described elsewhere (Villanova et al. 1998). The patient
is a boy with hydrocephalus and cobblestone lissencephaly with agyria and agenesis of the corpus callosum. Hypoplasia of the cerebellar vermis was seen in
the posterior cranial fossa. Ophthalmological examination showed buphthalmos, retinal dysplasia, and
lens opacities. Serum CK levels were 13,500 U/liter.
Patient 6 is an affected girl from unrelated Dutch par-
1036
ents. She has hydrocephalus, cobblestone lissencephaly,
cerebellar atrophy, bilateral microphthalmia, and frontal
bossing. A previous pregnancy was terminated because
of exencephaly.
The patients from family 7 are MZ twin girls from
an unaffected, nonconsanguineous Austrian couple.
Both present with encephalocele, pachygyria/agyria on
magnetic-resonance imaging, large ventricles, cerebellar vermis hypoplasia/aplasia, corpus callosum hypoplasia, markedly raised CK, and cataract. One child
has bilateral microtia.
Am. J. Hum. Genet. 71:1033–1043, 2002
antibody (mAb) raised against a-dystroglycan (VIA4-1;
Upstate Biotechnology), mAb anti-b-dystroglycan (clone
8D5; Novocastra), mAb anti–laminina-a2 (MAB 1922;
Chemicon), and mAb anti–laminin-a2 directed against
the 300-kDa subunit (4H8-2; Alexis). Sections were preincubated with PBS containing 5% BSA at pH 7.4, then
were incubated with the primary antibodies overnight at
4⬚C, and subsequently were incubated with the secondary
antibodies for 1 h at room temperature. Control sections
with omission of the first antibody were run in parallel.
Sections were examined and photographed on a Zeiss
Axioplan fluorescence microscope.
Linkage Analysis
Genomic DNA was extracted, by standard methods,
from peripheral-blood lymphocytes. In a first attempt to
clone the WWS gene, we performed a 10-cM-average
genomewide screening on 10 consanguineous families
with WWS. For the POMT1 linkage analysis, we tested
three microsatellite markers by PCR: D9S260, D9S64,
and D9S1793. In brief, amplification was performed in
a total volume of 20 ml (containing 40 ng of genomic
DNA, 5 pmol of each primer, 1 U Taq DNA polymerase
[Invitrogen], 250 mM of each dNTP, 1.5 mM MgCl2 [or
1.25 mM, in the case of D9S64], 50 mM KCl, and 20
mM Tris-HCl [pH 8.4; pH 9.5 in the case of D9S64]).
PCR conditions were as follows: 1 cycle at 94⬚C for 4
min; 35 cycles at 94⬚C for 1 min, 55⬚C for 1 min, and
72⬚C for 1 min; and 1 cycle at 72⬚C for 4 min. Samples
were resolved on 8% polyacrylamide sequencing gels and
were developed by silver staining (Budowle et al. 1991).
Mutation Analysis
The 19 coding exons of the POMT1 gene were amplified using specific primers for the 5- and 3-flanking
intron sequences. The PCR conditions were optimized
using the PCR Optimizer Kit (Invitrogen). The primer
sequences and conditions are available on request. The
amplicons generated were purified from agarose gels and
were directly sequenced with the BigDye terminator kit
(Perkin Elmer Applied Biosystems). Sequences were analyzed on an ABI3700 capillary sequencer (Perkin Elmer
Applied Biosystems). Testing of the mutations and SNPs
in the normal population was performed by SSCP analysis or by restriction-enzyme analysis in the event of the
Gln303Stop (BfaI) and Gln385Stop (PvuII) mutations.
Immunohistochemistry
Skeletal-muscle biopsy specimens were shock-frozen in
isopentane cooled in liquid nitrogen. Immunostaining of
6–8 mm cryosections was performed as described elsewhere (Kano et al. 2002) by using the following primary
antibodies: affinity-purified polyclonal sheep antibodies
directed against a 20-amino-acid C-terminal sequence of
chick a-dystroglycan (Herrmann et al. 2000), monoclonal
Results
Genomewide Linkage Analysis
To unravel the genetic basis of WWS, we performed
homozygosity mapping in 10 consanguineous families
with WWS, using 10-cM-spaced genetic markers. In a
first approach, we based our search on a one-locus
premise. Some good candidate regions of common homozygosity were found, but further analysis denied the
existence of a unique locus compatible with all families
with WWS. Moreover, further studies based on a twoloci premise were negative and suggested the existence
of at least three WWS loci. The high number of candidate loci and the minute amounts of patient samples
available prompted us to combine the homozygosity
mapping with a candidate gene approach. We focused
on genes involved in glycosylation and investigated
whether O-mannosyl glycan synthesis might underlie
WWS. The first step of the synthesis of these glycans
is executed by O-mannosyltransferases. As inferred
from sequence homology with the O-mannosyltransferases Pmt1–Pmt7 from yeast (Strahl-Bolsinger et al.
1999), four putative human counterparts have been
described: POMT1, POMT2, SDF2, and SDF2L1 (Hamada et al. 1996; Jurado et al. 1999; Fukuda et al.
2001). The protein domains of POMT1 and POMT2
show a similar arrangement to the yeast Pmts, whereas
SDF2 and SDF2L1 lack the hypothetical catalytic domain located in the first 300 amino acids. We selected
markers flanking the POMT1 (D9S260 and D9S1793;
on 9q34) and POMT2 (D14S254 and D14S287; on
14q24) genes for homozygosity mapping in 15 consanguineous families. The results were especially
promising in the case of POMT1. To substantiate the
possible linkage at the POMT1 locus, we tested an
additional genetic marker, D9S64, which is located in
intron 2 of the POMT1 gene. Homozygosity was observed in patients from 5 of the 15 consanguineous
families at this locus; only 1 of these 5 families was
already included in the genomewide linkage analysis.
Human POMT1 encodes a putative O-mannosyl-
1037
Beltrán-Valero de Bernabé et al.: POMT1 Mutations in WWS
Table 1
Mutations and Polymorphisms of POMT1 in Families with WWS
HAPLOTYPES ASSOCIATED
FAMILY (ORIGIN)a CONSANGUINITY
1 (Turkey)
2 (Turkey)
3 (Turkey)
4 (Italy)
5a (Turkey)
5b (Turkey)
6a (Netherlands)
6b (Netherlands)
7 (Austria)
Yes
Yes
Yes
No
Yes
Yes
No
No
No
MUTATION
c.226GrA (G76R)
c.907CrT (Q303X)
c.907CrT (Q303X)
c.2110InsG (V703fs)
NM
NM
c.1283TrA (V428D)
c.2167InsG (G722fs)
c.1153CrT (Q385X)
WITH
POMT1 MUTATIONSb
D9S260 c.1-6 D9S64 AA251 c.942 c.1148⫹16 c.2244⫹41 D9S1793
186
182
182
184
186
182
182
182
182
T
T
T
G
T
T
T
T
T
105
107
107
111
105
105
105
103
105
R
R
R
Q
W
W
R
R
R
C
C
C
T
C
C
C
C
C
G
G
G
G
A
A
G
G
G
C
C
C
T
C
C
C
C
C
182
182
182/172
188
188
172
182/172
182/172
186/178
NOTE.—NM p no mutation found.
a
The chromosomes are identified by the pedigree code number, followed by “a” or “b” (which refer to each of the two haplotypes identified
in a heterozygous patient; no letter is added if the patient is homozygous).
b
Allelic associations of eight genetic polymorphisms (ordered from 5 to 3) for each of the families with linkage to POMT1 are shown.
D9S260, D9S64, and D9S1793 are dinucleotide repeats, and the others are novel SNPs. D9S64 is located in POMT1 intron2, D9S260 is 3.1
cM toward the centromere, and D9S1793 is 3.9 cM toward the telomere. Note the shared haplotype for families 2 and 3, which carry the
same nonsense mutation, Gln303Stop.
transferase, which has different isoforms as the result
of alternative splicing. POMT1 is ubiquitously expressed with peak levels in adult testis, in skeletal and
cardiac muscle, and in fetal brain (Jurado et al. 1999).
This expression profile perfectly matches the tissues
affected in WWS. In addition to muscle, eye, and brain
involvement, male patients with WWS often have testicular defects (Hung et al. 1998).
POMT1 Mutation Analysis
Primers were designed for amplification of each of
the 20 exons and flanking intron sequences of POMT1.
Mutation analysis was performed by direct sequencing
of the resulting amplicons from patients from the five
families with linkage.
Patient 1 has a homozygous transition, 226GrA,
predicting a Gly76Arg substitution, in exon 3 (fig. 2A).
Gly76 is conserved in POMT1 orthologues from different species and is located in the second transmembrane (TM2) domain of the protein (Jurado et al.
1999). The substitution of this hydrophobic amino acid
by a basic arginine is likely to disrupt the membrane
affinity of this TM2 domain. Patients 2 and 3 are homozygous for a transversion, 907CrT, that creates a
Gln303Stop mutation in exon 9 (fig. 2B). Patient 4 is
homozygous for a frameshift mutation, 2110InsG—
predicted to cause a replacement of 44 highly conserved
C-terminal amino acids by 26 irrelevant ones after
Val703—in exon 20 (fig. 2C). No mutation was detected in patient 5. The Mendelian segregation of the
mutations was confirmed by direct sequencing of family
members. Testing of the mutations in the normal population was performed by SSCP analysis or by restriction-enzyme digestion in the case of Gln303Stop. This
mutation generates a new BfaI restriction site in exon
9. None of the changes were found in 60 Dutch and
45 Turkish control individuals. During the mutation
analysis, several SNPs were identified in the coding region and flanking intronic sequences of the POMT1
gene. Using these SNPs and dinucleotide markers, we
could assign haplotypes for each of the WWS chromosomes. The comparison of WWS haplotypes revealed a shared haplotype and, therefore, a possible
common ancestor for the two Turkish families carrying
the Gln303Stop mutation (table 1). The other mutations are associated with different haplotypes, indicating
a different origin for each of the POMT1 mutations.
Having established the causative role that POMT1 gene
mutations play in most of the families with WWS and
linkage, we sequenced this gene in all available patients
with WWS, originating from 10 consanguineous and 15
nonconsanguineous additional families. We found mutations in two of these families. Patient 6 is a compound
heterozygote (fig. 2D). The first mutation is a transversion, 1283TrA, that causes the replacement of a highly
conserved valine by an aspartic acid. The second mutation, 2167InsG, creates a frameshift that is predicted to
remove the 25 amino acids following Gly722. Both of the
MZ twin girls affected in family 7 are homozygous for a
transition, 1153CrT, that gives rise to a stop codon at
Gln385. This mutation disrupts a PvuII restriction site
used to verify the absence of this mutation in the normal
population. Thus, in total, we have found POMT1 mutations in 6 of the 30 families with WWS; this finding
supports the notion, from the genomewide linkage analysis, that WWS is genetically heterogeneous. We next extended our mutational analysis to the other putative Omannosyltransferases: POMT2, SDF2, and SDF2L1.
1039
Beltrán-Valero de Bernabé et al.: POMT1 Mutations in WWS
Direct sequencing did not reveal causative mutations in
any of the families with WWS that were sequenced.
Immunological Studies
To investigate whether O-mannosylation is indeed
abnormal in patients with WWS who carried POMT1
mutations, we performed immunohistochemical analysis of the highly O-mannosylated protein a-dystroglycan (Herrmann et al. 2000; Kano et al. 2002) in WWS
muscle biopsies (fig. 3). Dystroglycan is synthesized as
a precursor that generates two polypeptides by posttranslational cleavage: a- and b-dystroglycan. b-Dystroglycan is attached to the cell membrane, whereas a-dystroglycan associates with it and, through its carbohydrate tails, establishes other interactions with the extracellular matrix proteins, such as laminin-a2, perlecan,
and agrin. We have used two different antibodies
raised against a-dystroglycan: the monoclonal antibody,
VIA4.1, which is thought to recognize the carbohydrate
epitope of a-dystroglycan specifically, and the polyclonal
antibody, paDAG, which is raised against the C-terminal
end of the protein (Herrmann et al. 2000). VIA4.1 shows
a complete lack of staining in muscle biopsies of two
independent patients with WWS who carry homozygous
POMT1 mutations. paDAG immunostudies in the same
patients show a marked reduction of a-dystroglycan
peptide in many fibers. Misprocessing of the dystroglycan precursor protein is unlikely to be the cause of this
reduction, since b-dystroglycan immunoreactivity is normal. These results suggest that, in muscle from these
patients with WWS, a-dystroglycan not only is present
at reduced levels but also is underglycosylated. The adystroglycan O-mannosyl glycans are critical for binding
of laminin-a2, which connects the sarcolemma with the
extracellular matrix (Brown et al. 1999). Thus, the disruption of this interaction may be sufficient by itself to
cause the muscular dystrophy in WWS.
Figure 2
Discussion
In the present article, we have established that mutations
in the POMT1 gene cause WWS. Of 30 patients tested,
only 6 (20%) show POMT1 mutations. The genetic heterogeneity of the disease was already evidenced by the
genomewide linkage analysis, which demonstrated the
existence of at least three WWS loci, and excluded linkage for the POMT1 locus in several families.
The autosomal recessive inheritance pattern of WWS
and the type of mutations identified in patients suggest
that the syndrome is caused by a loss-of-function mechanism. POMT1 is not the only human enzyme that is
able to catalyze the first step in the O-mannosyl glycan
synthesis. Another putative O-mannosyltransferase is
POMT2, which shows an expression pattern in adults
that overlaps with that from POMT1; both POMT1 and
POMT2 show high expression levels in skeletal and cardiac muscle (GenBank accession number AF105020). Despite its similar function and expression profile, POMT2
is apparently unable to compensate for POMT1 loss-offunction mutations. This could be due to differences in
their expression patterns during embryonic development
or because of a differential arrangement of proteins that
are modified by these two mannosyltranferases. We have
not found any POMT2 mutation in 30 unrelated patients
with WWS.
The disruptive effect of the POMT1 mutations is likely
mediated through the reduced or absent O-mannosylation of target proteins, which affects their normal function in development. The implication, in MEB and WWS,
of two glycosyltransferases involved in O-mannose glycan synthesis clearly shows the relevance of O-mannosylation for eye, muscle, and brain development. All
patients carrying POMT1 mutations show the typical
features of WWS (fig. 1), as do other patients in whom
no mutations were detected. A previous reported linkage
analysis (Cormand et al. 2001), together with the genomewide linkage data from the present study, allows us
to exclude the MEB and FCMD loci in 8 of 10 families
POMT1 gene mutations in families with WWS. For each of the families with POMT1 mutations, the pedigrees are shown, as
well as two chromatograms, corresponding to a patient and a control DNA from the normal population. Patients for whom sequencing results
are shown are marked by an arrow. A, Family 1. The patient is homozygous for a 226GrA mutation (Gly76Arg). A multiple-sequence alignment
of the POMT1 peptide sequence from different species around glycine 76 is shown. * p not known. B, Families 2 and 3, carrying the same
homozygous mutation, 907CrT (Gln303Stop). This mutation creates a new BfaI site, which was used to analyze the segregation of this mutation
in family 3. The 180-bp amplicon is digested into 23- and 157-bp products for the wild-type allele, and the mutant allele is digested into
products of 110, 47, and 23 bp. Undigested (⫺) and digested (⫹) products were loaded for the control. C, Sequence analysis of family 4 reveals
homozygosity of 2110InsG (Val703fs). The bottom portion shows the comparison of the C-terminal peptide sequence from different species,
as well as the predicted protein encoded by the 2110InsG and 2167InsG (Gly722fs) alleles. The mutation is marked by an arrow. * p predicted
stop codon. D, Patient with WWS from family 6, who is a compound heterozygote for the POMT1 mutations 1283TrA (Val428Asp) and
2167InsG (Gly722fs). * p not known. E, MZ twins from family 7, who are homozygous for the 1153CrT (Gln385Stop) mutation. This
mutation disrupts a PvuII site, which was used to study the segregation in the family. The 178-bp amplicon is digested into 78- and 100-bp
products for the wild-type allele, and into 49-, 51-, and 78-bp products for the mutant allele.
1040
Am. J. Hum. Genet. 71:1033–1043, 2002
Figure 3
Immunofluorescence staining of skeletal muscle from a control individual and from patients with WWS from families 2 and 4.
Note the normal expression of laminin-a2 and b-dystroglycan in WWS. Variable staining intensities were observed with anti-a-dystroglycan
antibody specific for the C-terminal peptide sequence, with most muscle fibers showing a marked reduction in a-dystroglycan staining. No adystroglycan staining was observed in patient muscle by using an antibody thought to recognize the O-glycan motif.
with WWS that were investigated. Given that the WWS
and MEB genes are both involved in O-mannosylation
and given the genetic heterogeneity shown for these related syndromes, it is very likely that other proteins involved in this type of glycosylation underlie the as-yetunexplained patients.
All of the O-mannose–linked glycans identified to date
contain the motif Galb1 4GlcNAcb1 2Man-O-Ser/Thr
(O-mannose–linked core), to which other sugar residues
can be sequentially added (fig. 4). Most of the resultant
variants are brain specific, and only one such motif has
been identified in both brain and muscle (Siaa2 3Galb1
4GlcNAcb1 2Man-O-Ser/Thr; fig. 4). Both POMGnT1
and POMT1 are involved in the synthesis of the O-
Beltrán-Valero de Bernabé et al.: POMT1 Mutations in WWS
1041
Figure 4
Representation of the only O-mannose–linked glycan motif found in brain, muscle, and peripheral nerve. The big circle represents
the protein, to which several sugar residues are added in a sequential manner by the catalytic activity of specific glycosyltransferases. POMT1
catalyzes the addition of the first residue (mannose), whereas POMGnT1 catalyzes the second step, the linkage of N-acetylglucosamine in b1,2. B4GALTs (Amado et al. 1999) and ST3GALs (Tsuji 1996) are the families of enzymes that are able to catalyze the third and fourth steps,
respectively, of this O-glycan synthesis. The specific enzymes that catalyze these steps have not been assigned yet.
mannose–linked core. Since POMT1 catalyzes the first
step and POMGnT1 catalyzes the second step in O-mannose glycosylation (fig. 4), one might expect the phenotype associated with POMT1 mutations to be more
severe than that associated with POMGnT1 mutations.
One might therefore speculate that mutation of glycosyltransferases that catalyze subsequent steps gives rise
to a phenotype that is less severe than MEB/WWS. However, the level of redundancy at each of the glycosylation
steps is not known, and, therefore, it can also be envisaged that disruption of an enzyme that catalyzes a later
step will give rise to an equally severe phenotype.
The O-glycans attached to a-dystroglycan play a
role in molecular recognition (Brown et al. 1999).
The immunochemistry results shown in the present
article illustrate a secondary reduction in the protein
levels of a-dystroglycan. Secondary reductions of dystrophin-glycoprotein complex (DGC) components,
in which a-dystroglycan is included, have been described in other muscular dystrophies (Ohlendieck
and Campbell 1991; Barresi et al. 1997). The stability
of each component of the DGC depends on the correct interaction with its neighboring proteins, and the
absence of one of the components of the DGC often
causes a reduction of its interacting proteins as well.
Most likely the depletion of the O-mannosyl glycans
prevents the normal interaction between a-dystroglycan and laminin-a2, and this disturbance reduces
the a-dystroglycan stability. In agreement with this,
several patients with WWS who have been described
in the literature showed reduced levels of the a-dystroglycan–interacting muscle proteins laminin-a2 (Kanoff
et al. 1998), laminin-b2, and adhalin (Wewer et al.
1995).
Hypoglycosylation of a-dystroglycan may well be sufficient to explain the muscular dystrophy that is observed
in patients with WWS. O-glycans are required for the
attachment of a-dystroglycan to the extracellular matrix,
which is crucial for the correct contraction of the muscle.
Apart from WWS, MEB, and FCMD, mutations in two
other glycosyltransferases—Large, in mice, and FKRP, in
human—cause hypoglycosylation of a-dystroglycan and
muscular dystrophy (Brockington et al. 2001a; Grewal
et al. 2001). FKRP and Large mutants do not present
severe brain symptoms.
The brain and eye phenotypes in WWS and MEB
probably involve defective glycosylation of proteins
other than a-dystroglycan, since chimeric mice deficient
in dystroglycan develop muscular dystrophy but do not
display a brain or eye phenotype (Cote et al. 1999). In
brain, other possible O-mannosylation targets that may
be responsible for the WWS symptoms include reelin
and tenascin-J1 (Lochter et al. 1991; Hirotsune et al.
1995). Both proteins are diffusible factors that show a
gradient distribution in brain development. Reelin is a
glycoprotein secreted by Cajal-Retzius cells and other
pioneer neurons and is implicated in lissencephaly with
cerebellar hypoplasia (Hong et al. 2000). Tenascin-J1
is a highly O-mannosylated glycoprotein that shows repellent activity on neurite extension in neuroblast culture (Wing et al. 1992).
The neuronal overmigration and the invasion of the
subarachnoid space by the migrating neurons in the
WWS neocortex (figs. 1E and 1F) resembles the situation present in laminin-a6 and myristoylated alaninerich protein C kinase substrate (MARCKS) knockout
mice (Stumpo et al. 1995; Georges-Labouesse et al.
1998). In the MARCKS knockout, it has been hypothesized that the disturbance of structural components of
the basal lamina is responsible for the neuronal overmigration, whereas, in laminin-a6, the recognition process between the matrix and the migrating neurons is
disrupted. Similarly failed interactions with extracellular matrix components may underlie the severe neuronal overmigration that causes cobblestone lissencephaly in WWS. It is interesting that, whereas integrin-
1042
a3 and integrin-a6 have been implicated in neocortex
lamination, integrin-a5 and integrin-a7 are involved in
muscle development (Mayer et al. 1997; Georges-Labouesse et al. 1998; Taverna et al. 1998; Anton et al.
1999). It is tempting to speculate that the a-integrins
may also be targets for O-mannosylation.
Note added in proof.—While the present article was
in press, two papers were published that are relevant to
the pathogenesis of WWS. The first article (Michele et
al. 2002) shows that, in patients with MEB and FCMD,
a-dystroglycan is hypoglycosylated and no longer interacts with proteins from the extracellular matrix. In the
second article (Moore et al. 2002), it is demonstrated
that the brain-specific disruption of a-dystroglycan during mouse embryonic development causes neuronal overmigration and fusion of cerebral hemispheres, resembling WWS brain abnormalities. These papers indicate
that the disrupted glycosylation of a-dystroglycan makes
a major contribution to the CNS abnormalities in WWS.
Acknowledgments
We thank Drs. Rudolf Puttinger, Yvonne Hilhorst, Christine
de Die, Koen DeVriendt, and Ernie Bongers, for contribution
of family material for this study; Dr. Henk ter Laak, for immunohistochemical analysis of a-dystroglycan; and Dr. Luis
Alberto Pérez Jurado, for his previous work on the POMT1
gene. The genomewide linkage analysis of some of the families
was performed at the Centre de Genotypage, Paris. This work
was supported by grants from the Dutch Foundation for Scientific Research (NWO, 903-42-190), the Prinses Beatrix Fund
(MAR00-117), and the KNAW–van Leersum Fund. T.V. and
L.M. were supported by E.U. grant QLRT-1999-00870, and
T.V. was also supported by the Deutsche Forschungs Gemeinschaft (Str498/3-1); C.A.W. was supported by the National
Institutes of Health (RO1 NS 35129).
Electronic-Database Information
Accession numbers and URLs for data presented herein are
as follows:
GenBank, http://www.ncbi.nlm.nih.gov/Genbank/ (for POMT2
[accession number AF105020])
Online Mendelian Inheritance in Man (OMIM), http://www
.ncbi.nlm.nih.gov/Omim/ (for WWS [MIM 236670], MEB
[MIM 253280], and FCMD [MIM 253800])
References
Amado M, Almeida R, Schwientek T, Clausen H (1999) Identification and characterization of large galactosyltransferase
gene families: galactosyltransferases for all functions. Biochim
Biophys Acta 1473:35–53
Anton ES, Kreidberg JA, Rakic P (1999) Distinct functions of
a3 and aV integrin receptors in neuronal migration and laminar organization of the cerebral cortex. Neuron 22:277–289
Aravind L, Koonin EV (1999) The fukutin protein family—
Am. J. Hum. Genet. 71:1033–1043, 2002
predicted enzymes modifying cell-surface molecules. Curr
Biol 9:R836–R837
Barresi R, Confalonieri V, Lanfossi M, Di Blasi C, Torchiana
E, Mantegazza R, Jarre L, Nardocci N, Boffi P, Tezzon F,
Pini A, Cornelio F, Mora M, Morandi L (1997) Concomitant
deficiency of b- and g-sarcoglycans in 20 a-sarcoglycan
(adhalin)-deficient patients: immunohistochemical analysis
and clinical aspects. Acta Neuropathol (Berl) 94:28–35
Brockington M, Blake DJ, Prandini P, Brown SC, Torelli S, Benson MA, Ponting CP, Estournet B, Romero NB, Mercuri E,
Voit T, Sewry CA, Guicheney P, Muntoni F (2001a) Mutations
in the fukutin-related protein gene (FKRP) cause a form of
congenital muscular dystrophy with secondary laminin a2
deficiency and abnormal glycosylation of a-dystroglycan.
Am J Hum Genet 69:1198–1209
Brockington M, Yuva Y, Prandini P, Brown SC, Torelli S, Benson MA, Herrmann R, Anderson LV, Bashir R, Burgunder
JM, Fallet S, Romero N, Fardeau M, Straub V, Storey G,
Pollitt C, Richard I, Sewry CA, Bushby K, Voit T, Blake DJ,
Muntoni F (2001b) Mutations in the fukutin-related protein gene (FKRP) identify limb girdle muscular dystrophy
2I as a milder allelic variant of congenital muscular dystrophy MDC1C. Hum Mol Genet 10:2851–2859
Brown SC, Fassati A, Popplewell L, Page AM, Henry MD,
Campbell KP, Dickson G (1999) Dystrophic phenotype induced in vitro by antibody blockade of muscle a-dystroglycan-laminin interaction. J Cell Sci 112 (Pt 2):209–216
Budowle B, Chakraborty R, Giusti AM, Eisenberg AJ, Allen
RC (1991) Analysis of the VNTR locus D1S80 by the PCR
followed by high-resolution PAGE. Am J Hum Genet 48:
137–144
Chai W, Yuen CT, Kogelberg H, Carruthers RA, Margolis RU,
Feizi T, Lawson AM (1999) High prevalence of 2-mono- and
2,6-di-substituted manol-terminating sequences among O-glycans released from brain glycopeptides by reductive alkaline
hydrolysis. Eur J Biochem 263:879–888
Chiba A, Matsumura K, Yamada H, Inazu T, Shimizu T, Kusunoki S, Kanazawa I, Kobata A, Endo T (1997) Structures
of sialylated O-linked oligosaccharides of bovine peripheral
nerve alpha-dystroglycan: the role of a novel O-mannosyltype oligosaccharide in the binding of a-dystroglycan with
laminin. J Biol Chem 272:2156–2162
Cormand B, Pihko H, Bayes M, Valanne L, Santavuori P,
Talim B, Gershoni-Baruch R, Ahmad A, van Bokhoven H,
Brunner HG, Voit T, Topaloglu H, Dobyns WB, Lehesjoki
AE (2001) Clinical and genetic distinction between WalkerWarburg syndrome and muscle-eye-brain disease. Neurology 56:1059–1069
Cote PD, Moukhles H, Lindenbaum M, Carbonetto S (1999)
Chimaeric mice deficient in dystroglycans develop muscular
dystrophy and have disrupted myoneural synapses. Nat Genet 23:338–342
Dobyns WB, Pagon RA, Armstrong D, Curry CJ, Greenberg
F, Grix A, Holmes LB, Laxova R, Michels VV, Robinow M
(1989) Diagnostic criteria for Walker-Warburg syndrome.
Am J Med Genet 32:195–210
Endo T (1999) O-mannosyl glycans in mammals. Biochim Biophys Acta 1473:237–246
Fukuda S, Sumii M, Masuda Y, Takahashi M, Koike N, Teishima
J, Yasumoto H, Itamoto T, Asahara T, Dohi K, Kamiya K
(2001) Murine and human SDF2L1 is an endoplasmic retic-
Beltrán-Valero de Bernabé et al.: POMT1 Mutations in WWS
ulum stress-inducible gene and encodes a new member of the
Pmt/rt protein family. Biochem Biophys Res Commun 280:
407–414
Georges-Labouesse E, Mark M, Messaddeq N, Gansmüller A
(1998) Essential role of a6 integrins in cortical and retinal
lamination. Curr Biol 8:983–986
Grewal PK, Holzfeind PJ, Bittner RE, Hewitt JE (2001) Mutant
glycosyltransferase and altered glycosylation of a-dystroglycan in the myodystrophy mouse. Nat Genet 28:151–154
Hamada T, Tashiro K, Tada H, Inazawa J, Shirozu M, Shibahara K, Nakamura T, Martina N, Nakano T, Honjo T
(1996) Isolation and characterization of a novel secretory
protein, stromal cell-derived factor-2 (SDF-2) using the signal sequence trap method. Gene 176:211–214
Herrmann R, Straub V, Blank M, Kutzick C, Franke N, Jacob
EN, Lenard HG, Kroger S, Voit T (2000) Dissociation of
the dystroglycan complex in caveolin-3-deficient limb girdle
muscular dystrophy. Hum Mol Genet 9:2335–2340
Hirotsune S, Takahara T, Sasaki N, Hirose K, Yoshiki A,
Ohashi T, Kusakabe M, Murakami Y, Muramatsu M,
Watanabe S, Nakao K, Katsuki M, Hayashizaki Y (1995)
The reeler gene encodes a protein with an EGF-like motif
expressed by pioneer neurons. Nat Genet 10:77–83
Hong SE, Shugart YY, Huang DT, Shahwan SA, Grant PE,
Hourihane JO, Martin ND, Walsh CA (2000) Autosomal
recessive lissencephaly with cerebellar hypoplasia is associated with human RELN mutations. Nat Genet 26:93–96
Hung NA, Silver MM, Chitayat D, Provias J, Toi A, Jay V,
Becker LE (1998) Gonadoblastoid testicular dysplasia in
Walker-Warburg syndrome. Pediatr Dev Pathol 1:393–404
Jurado LA, Coloma A, Cruces J (1999) Identification of a
human homolog of the Drosophila rotated abdomen gene
(POMT1) encoding a putative protein O-mannosyl-transferase, and assignment to human chromosome 9q34.1.
Genomics 58:171–180
Kano H, Kobayashi K, Herrmann R, Tachikawa M, Manya
H, Nishino I, Nonaka I, Straub V, Talim B, Voit T, Topaloglu
H, Endo T, Yoshikawa H, Toda T (2002) Deficiency of adystroglycan in muscle-eye-brain disease. Biochem Biophys
Res Commun 291:1283–1286
Kanoff RJ, Curless RG, Petito C, Falcone S, Siatkowski RM,
Pegoraro E (1998) Walker-Warburg syndrome: neurologic
features and muscle membrane structure. Pediatr Neurol 18:
76–80
Kobata A (1992) Structures and functions of the sugar chains
of glycoproteins. Eur J Biochem 209:483–501
Kobayashi K, Nakahori Y, Miyake M, Matsumura K, Kondo-Iida E, Nomura Y, Segawa M, Yoshioka M, Saito K,
Osawa M, Hamano K, Sakakihara Y, Nonaka I, Nakagome Y, Kanazawa I, Nakamura Y, Tokunaga K, Toda
T (1998) An ancient retrotransposal insertion causes
Fukuyama-type congenital muscular dystrophy. Nature
394:388–392
Lammens M (2000) Neuronal migration disorders in man. Eur
J Morphol 38:327–333
Lochter A, Vaughan L, Kaplony A, Prochiantz A, Schachner
M, Faissner A (1991) J1/tenascin in substrate-bound and
soluble form displays contrary effects on neurite outgrowth.
J Cell Biol 113:1159–1171
Marin-Padilla M (1978) Dual origin of the mammalian neocortex and evolution of the cortical plate. Anat Embryol
(Berl) 152:109–126
1043
Mayer U, Saher G, Fassler R, Bornemann A, Echtermeyer F,
von der Mark H, Miosge N, Poschl E, von der Mark K
(1997) Absence of integrin alpha 7 causes a novel form of
muscular dystrophy. Nat Genet 17:318–323
Michele DE, Barresi R, Kanagawa M, Saito F, Cohn RD, Satz
JS, Dollar J, Nishino I, Kelley RI, Somer H, Straub V, Mathews KD, Moore SA, Campbell KP (2002) Post-translational disruption of dystroglycan-ligand interactions in congenital muscular dystrophies. Nature 418:417–422
Moore SA, Saito F, Chen J, Michele DE, Henry MD, Messing
A, Cohn RD, Ross-Barta SE, Westra S, Williamson RA,
Hoshi T, Campbell KP (2002) Deletion of brain dystroglycan
recapitulates aspects of congenital muscular dystrophy. Nature 418:422–425
Ohlendieck K, Campbell KP (1991) Dystrophin-associated
proteins are greatly reduced in skeletal muscle from mdx
mice. J Cell Biol 115:1685–1694
Sasaki T, Yamada H, Matsumura K, Shimizu T, Kobata A,
Endo T (1998) Detection of O-mannosyl glycans in rabbit
skeletal muscle alpha-dystroglycan. Biochim Biophys Acta
1425:599–606
Smalheiser NR, Haslam SM, Sutton-Smith M, Morris HR, Dell
A (1998) Structural analysis of sequences O-linked to mannose reveals a novel Lewis X structure in cranin (dystroglycan)
purified from sheep brain. J Biol Chem 273:23698–23703
Strahl-Bolsinger S, Gentzsch M, Tanner W (1999) Protein Omannosylation. Biochim Biophys Acta 1426:297–307
Stumpo DJ, Bock CB, Tuttle JS, Blackshear PJ (1995) MARCKS
deficiency in mice leads to abnormal brain development and
perinatal death. Proc Natl Acad Sci USA 92:944–948
Taverna D, Disatnik MH, Rayburn H, Bronson RT, Yang J,
Rando TA, Hynes RO (1998) Dystrophic muscle in mice chimeric for expression of a5 integrin. J Cell Biol 143:849–859
Tsuji S (1996) Molecular cloning and functional analysis of
sialyltransferases. J Biochem (Tokyo) 120:1–13
Varki A (1993) Biological roles of oligosaccharides: all of the
theories are correct. Glycobiology 3:97–130
Villanova M, Sabatelli P, He Y, Malandrini A, Petrini S, Maraldi NM, Merlini L (1998) Immunofluorescence study of a
muscle biopsy from a 1-year-old patient with Walker-Warburg syndrome. Acta Neuropathol (Berl) 96:651–654
Wewer UM, Durkin ME, Zhang X, Laursen H, Nielsen NH,
Towfighi J, Engvall E, Albrechtsen R (1995) Laminin beta 2
chain and adhalin deficiency in the skeletal muscle of WalkerWarburg syndrome (cerebro-ocular dysplasia-muscular dystrophy). Neurology 45:2099–2101
Wing DR, Rademacher TW, Schmitz B, Schachner M, Dwek
RA (1992) Comparative glycosylation in neural adhesion
molecules. Biochem Soc Trans 20:386–390
Yoshida A, Kobayashi K, Manya H, Taniguchi K, Kano H,
Mizuno M, Inazu T, Mitsuhashi H, Takahashi S, Takeuchi
M, Herrmann R, Straub V, Talim B, Voit T, Topaloglu H,
Toda T, Endo T (2001) Muscular dystrophy and neuronal
migration disorder caused by mutations in a glycosyltransferase, POMGnT1. Dev Cell 1:717–724
Yuen CT, Chai W, Loveless RW, Lawson AM, Margolis RU,
Feizi T (1997) Brain contains HNK-1 immunoreactive Oglycans of the sulfoglucuronyl lactosamine series that terminate in 2-linked or 2,6-linked hexose (mannose). J Biol
Chem 272:8924–8931